History of smoking and future of lung cancer mortality: A cross-national comparison.

Abstract

e16503 Background: The association of smoking and lung cancer mortality has been well established for over 50 years. Smoking patterns differ by age, sex, year, and country reflected in public health campaigns, taxation, and social stigma. Historically, smoking has been more prevalent among men than women, but lately it has declined more rapidly among men. We forecast lung cancer mortality over the next decades in two countries with different smoking patterns: In Norway, smoking rates declined slowly for men and remained stagnant for women, whereas in the US, smoking has declined more consistently, especially for men, but lately also among women. Methods: US and Norwegian death certificates and population counts (1969-2007) and historical smoking prevalence by age, sex, and year (1961-2007) were used. Bayesian hierarchical forecasting modeling was employed to forecast lung cancer mortality by age, sex, and year, taking into account known demographic patterns and smoking prevalence 25-years earlier in age and year. Results: Smoking levels were historically higher for men in both countries. The pace of decline was fastest for US men, next US women and Norwegian males, and finally relatively stagnant for Norwegian females. Strong time-lagged correlation of age-specific lung cancer mortality and smoking prevalence 25-years earlier in age and year. Between 2007 and 2032, we forecast the age-standardized crude death rate change from 80 to 53 for US males, 52 to 34 for US females, 58 to 74 for Norwegian males, and 34 to 91 for Norwegian females (lung cancer deaths per 100,000 person-years). Conclusions: The burden of lung cancer mortality is shifting to females, and there is thus a greater need for anti-smoking and lung cancer screening initiatives directed towards women in both the US and Norway. Whereas the total lung cancer burden will decline in the US in the near future, our forecast predicts that this burden will increase in Norway. This population-level mortality forecast may be useful for setting research priorities and allocating public health expenditures.